Effects of virtualization and cloud computing on data center networks

10
Figure 5: vSwitches do a good job of efficiently routing internal VM-to-VM traffic.
However, there are some limitations to a vSwitch:
• It moves the control point for networking infrastructure into the domain of the server administrator.
This management stack is typically a component of the server-based hypervisor tool aimed at system
and virtualization administrators. As such, vSwitch management generally does not integrate with
existing external physical network policy and management tools. This usually means two different
teams (with different processes) manage the physical network and the virtual network, even though
the management tasks and functionality overlap.
• It consumes valuable CPU bandwidth. The higher the traffic load, the greater the number of CPU
cycles required to move traffic through the vSwitch. This reduces the ability to support larger
numbers of VMs in a physical server.
• It lacks network-based visibility. A vSwitch does not have standard network monitoring capabilities
such as flow analysis, advanced statistics, and remote diagnostics of external network switches.
When network outages or problems occur, identifying the root cause can be difficult in a virtualized
server environment.
EVB architectures—VEPA and VEB
To solve some of these management drawbacks, HP is working with other vendors in the IEEE 802.1
Work Group to develop the Edge Virtual Bridging (EVB) standard. The EVB standard uses Virtual
Ethernet Port Aggregator (VEPA) technology as its foundation. VEPA is a way for virtual switches to
send all traffic and forwarding decisions to the adjacent physical switch (Figure 6). This removes the
burden of VM forwarding decisions and network operations from the host CPU. And it leverages the
existing management capabilities in the access-layer switches.